GPT CoPilot

Analyzes selected code blocks in the editor and generates natural language explanations using OpenAI's GPT-3 API. The extension captures the highlighted text through VS Code's selection API, sends it to OpenAI with a system prompt optimized for code explanation, and streams or returns the response to the Output panel. Works with any language VS Code syntax-highlights, leveraging GPT-3's multi-language code understanding without language-specific parsing.

Processes an entire file's content through OpenAI's GPT-3 API to generate comprehensive documentation or explanations. Unlike single-selection explanation, this capability reads the full file buffer via VS Code's document API and sends the complete source to GPT-3 with a documentation-focused prompt, returning structured or narrative documentation to the Output panel. Useful for generating module-level docstrings, README sections, or API documentation from source code.

Operates on a freemium model where the extension itself is free, but users pay OpenAI directly for API usage via their own API key. The extension has no built-in usage limits, quotas, or metering — all costs are incurred by the user based on their OpenAI API consumption. Free tier users can use the extension unlimited times as long as they have API credits; paid tiers are not required for the extension itself, only for OpenAI API access.

Accepts arbitrary natural language prompts from users and generates code snippets or completions using OpenAI's GPT-3 API. Users input prompts via the command palette or context menu, the extension sends the prompt to GPT-3 with optional context (current file, selection, or standalone), and returns generated code to the Output panel or clipboard. Supports concept elaboration and code generation without requiring highlighted code as input.

Allows users to specify which OpenAI GPT-3 model variant to use via VS Code settings (e.g., text-davinci-003, gpt-3.5-turbo). The extension reads the `gpt-copilot.model` configuration value at runtime and passes it to the OpenAI API request, enabling users to trade off cost, speed, and quality without modifying extension code. Supports any model available through the user's OpenAI API account.

Provides a configurable `gpt-copilot.maxTokens` setting that controls the maximum length of GPT-3 responses. The extension passes this value to the OpenAI API's `max_tokens` parameter, allowing users to constrain response length for cost control or conciseness. Shorter limits reduce API costs and latency; longer limits enable more detailed explanations or code generation.

Offers a configurable `gpt-copilot.temperature` setting (0-1 range) that controls the randomness and creativity of GPT-3 responses. Lower values (near 0) produce deterministic, focused explanations; higher values (near 1) produce more creative and varied responses. The extension passes this value to the OpenAI API's `temperature` parameter, enabling users to tune response behavior for different use cases.

Manages OpenAI API key storage securely using VS Code's built-in `SecretStorage API`, which encrypts credentials at rest and prevents exposure in plaintext configuration files. Users configure their API key via the `GPT - Setup` command in the command palette, which prompts for the key and stores it securely. The extension retrieves the key at runtime for API authentication without exposing it in settings files or logs.

Provides right-click context menu integration in VS Code's editor, allowing users to trigger code explanation, documentation, or querying operations without using the command palette. The extension registers context menu items that operate on the current selection or file, passing the selected code or file content to GPT-3 for processing. Specific menu items are undocumented but likely include 'Explain Code', 'Generate Documentation', or similar actions.

Displays all GPT-3 responses in VS Code's built-in Output panel under a 'GPT CoPilot' channel, providing a persistent, scrollable view of all explanations, documentation, and generated code. Results are appended to the output channel with timestamps or separators, allowing users to review history and compare multiple responses without switching windows. The Output panel integrates seamlessly with VS Code's UI and supports text selection, copying, and clearing.

Supports code explanation and generation across any programming language that VS Code syntax-highlights, leveraging GPT-3's broad language knowledge without language-specific parsing or AST analysis. The extension treats code as plain text and sends it to GPT-3, which understands syntax and semantics across Python, JavaScript, Java, C++, Go, Rust, and 50+ other languages. No language detection or validation is performed — GPT-3 infers language from syntax.